A graph-based estimator of the number of clusters
ESAIM: Probability and Statistics, Tome 11 (2007), pp. 272-280

Voir la notice de l'article provenant de la source Numdam

Assessing the number of clusters of a statistical population is one of the essential issues of unsupervised learning. Given n independent observations X 1 ,,X n drawn from an unknown multivariate probability density f, we propose a new approach to estimate the number of connected components, or clusters, of the t-level set (t)={x:f(x)t}. The basic idea is to form a rough skeleton of the set (t) using any preliminary estimator of f, and to count the number of connected components of the resulting graph. Under mild analytic conditions on f, and using tools from differential geometry, we establish the consistency of our method.

DOI : 10.1051/ps:2007019
Classification : 62G05, 62G20
Keywords: cluster analysis, connected component, level set, graph, tubular neighborhood
@article{PS_2007__11__272_0,
     author = {Biau, G\'erard and Cadre, Beno{\^\i}t and Pelletier, Bruno},
     title = {A graph-based estimator of the number of clusters},
     journal = {ESAIM: Probability and Statistics},
     pages = {272--280},
     publisher = {EDP-Sciences},
     volume = {11},
     year = {2007},
     doi = {10.1051/ps:2007019},
     mrnumber = {2320821},
     zbl = {1187.62114},
     language = {en},
     url = {http://geodesic.mathdoc.fr/articles/10.1051/ps:2007019/}
}
TY  - JOUR
AU  - Biau, Gérard
AU  - Cadre, Benoît
AU  - Pelletier, Bruno
TI  - A graph-based estimator of the number of clusters
JO  - ESAIM: Probability and Statistics
PY  - 2007
SP  - 272
EP  - 280
VL  - 11
PB  - EDP-Sciences
UR  - http://geodesic.mathdoc.fr/articles/10.1051/ps:2007019/
DO  - 10.1051/ps:2007019
LA  - en
ID  - PS_2007__11__272_0
ER  - 
%0 Journal Article
%A Biau, Gérard
%A Cadre, Benoît
%A Pelletier, Bruno
%T A graph-based estimator of the number of clusters
%J ESAIM: Probability and Statistics
%D 2007
%P 272-280
%V 11
%I EDP-Sciences
%U http://geodesic.mathdoc.fr/articles/10.1051/ps:2007019/
%R 10.1051/ps:2007019
%G en
%F PS_2007__11__272_0
Biau, Gérard; Cadre, Benoît; Pelletier, Bruno. A graph-based estimator of the number of clusters. ESAIM: Probability and Statistics, Tome 11 (2007), pp. 272-280. doi: 10.1051/ps:2007019

Cité par Sources :